Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

MA29

N.A. Railroads have traditionally scheduled trains with little respect to their interactions with each other. This is particularly true for unit trains and ad hoc trains. To increase network capacity, train performance predictability, and general network fluidity, railroads should change their approach to tactical train scheduling to take into account these interactions. Adding dispatch optimizers is not enough as a metering process is required to ensure that the plan is achievable and the number of trains the dispatcher is to handle will result in reliable operations. 2 - Interaction Between Yard and Mainline Capacity in Railway Network Performance Tyler Dick, U. of Illinois at Urbana-Champaign, 1241 Newmark Lab MC-250, 205 N. Mathews Avenue, Urbana, IL, 61801, United States Mainline, yard and terminal performance all contribute to the quality of railway service. Mainlines and yards are symbiotic; a disruption to one operation quickly manifests in reduced performance of the other. Although routinely observed, there has been little academic study of the capacity interactions between mainlines and yards. This presentation provides a high-level overview of the “network efficiency cycle linking mainline and yard performance, and research on the capacity impact of propagating schedule flexibility from yard-to-yard over connecting mainlines. 3 - Increasing Network Capacity through Use of Dynamic Car Scheduling Clark Cheng, Sr. Director Operations Res & Chief Data Scientist, Norfolk Southern Corporation, 1200 Peachtree Street NE, Mail Stop 171, Atlanta, GA, 30309, United States, Edward Lin, Gunnar Feldmann Car scheduling is the key operating system for freight railroads to classify railcars and generate trip plans. With dynamic car scheduling, the classification and trip plan for a shipment may vary in order to maximize capacity utilization while improve customer service. In this presentation, we will describe a dynamic car scheduling model developed at Norfolk Southern and how it can improve capacity utilization, operations efficiency, and customer service. 4 - Impact of Improved Maintenance Programs on Railroad Capacity David T. Hunt, Oliver Wyman, One University Square, Suite 100, Princeton, NJ, 08540, United States Railroads have progressed their maintenance practices from curative (fix it when it breaks), to systematic (fix it at predetermined intervals), to condition based (fix it based on inspections), and are now working on predictive approaches. This not only leads to safer operations and improved customer service, but has additional benefits in improved rail capacity. This presentation explores how improvements in rail maintenance have impacted capacity, and the corresponding economic consequences. n MA29 North Bldg 221B Joint Session TSL/Practice Curated: Game-theoretic Applications in Urban Transport Sponsored: TSL/Urban Transportation Sponsored Session Chair: Joseph Y. J. Chow, New York University, New York, NY, 10012, United States 1 - Highway Infrastructure Protection Against Sea Level Rise: Game Theoretical Approaches ILIA Papakonstantinou, NYU, Brooklyn, NY, 11209, United States This research investigates the differences between game theoretical approaches among decision makers protecting their infrastructure against sea level rise. We take into account the hydrodynamic interactions, transportation system adaptation to inundations due to sea level rise, and budget constraints. Both competitive and cooperative games are considered, where each decision maker aims to minimize traffic delays within its territory caused by the network disruptions due to sea level rise. A case study of the San Francisco Bay area for 0.5m sea level rise is used. The results of different games are compared for a range of budgeting scenarios. 2 - A Many-to-many Auction-based Multimodal Network Assignment Game Theodoros P. Pantelidis, New York University, Brooklyn, NY, United States Urban transportation networks are multi-dimensional complex systems that serve large numbers of travelers every day. These complex multimodal systems can be considered as multiple smaller networks which are privately owned by transit operators. When considering pricing and capacity decisions, operators take into account only volume flows in their own network and fail to capture the effects of multimodal trips. In this study, we propose an auction-based game theoretic approach, where travelers act as buyers and network edges as market products that belong to a set of sellers, in order to find the optimal allocation that maximizes joint profits.

n MA27 North Bldg 132B Analytics Pedagogy Sponsored: Education (INFORMED) Sponsored Session Chair: Matthew D. Dean, University of Southern Maine, Portland, ME, 04104, United States 1 - Concurrent Undergraduate and Graduate Analytics Program Development Lessons Learned Stephen Hill, University of North Carolina-Wilmington, 601 South College Road, Wilmington, NC, 28403-5611, United States Many universities have developed or are developing analytics programs. In this presentation, we will discuss the challenges associated with the simultaneous development of graduate and undergraduate analytics programs. We will share lessons learned and opportunities for improvement. 2 - Teaching Analytics Methods with Jupyter Notebooks and Python Nickolas K. Freeman, University of Alabama, 14485 Griffin St, Tuscaloosa, AL, 35405, United States This talk will discuss the use of Jupyter notebooks and the Python programming language for the delivery of analytics content in a classroom setting. Jupyter notebooks are capable of combining rich text elements with executable code, allowing students to execute and visualize analyses in real-time. A simple demonstration will illustrate how the technology may be used to facilitate the presentation of advanced analytics methods. Lessons learned from a recent course based on the technology will also be discussed. 3 - Using the Learning Assistant Model in an Undergraduate Business Analytics Course Matthew D. Dean, University of Southern Maine, School of Business, 113 Luther Bonney Hall, Portland, ME, 04104, United States We report the results of the first known implementation of the Learning Assistant Model in a business analytics course (or any business-related course). This pedagogical technique shows promise in helping students master the material better than those who took the course in a traditional lecture-based learning environment. Students perceptions of the Learning Assistant Model learning environment also improved from the beginning of the semester to the end. 4 - Teaching Experiences of an Elective Operations Research Course in a Business School Narasimhan Ravichandran, Professor, Indian Institute of Management Ahmadabad, Wing 2 D, Vastrapur, Ahmedabad, 380015, India I have been teaching a case based course on the applications of OR as an elective in the MBA program at IIMA. Given the changed profile of the participants and the job market compulsions there are some real challenges in teaching this course. In the last three editions we have deployed some innovative methods to deal with the challenges. These ideas are presented in this talk and I explore the options for the future editions of the course. 5 - Lecture Based Tutoring: A New Active Learning Teaching Technique Todd W. Easton, Kansas State University, 2037 Durland Hall, Manhattan, KS, 66506, United States This talk describes a new active learning teaching technique called lecture based tutoring. The premise of lecture based tutoring is to tutor students during lectures. The teacher calls a student by name and asks a question. If the student answers poorly, then the student is tutored to the correct answer. Between 20 and 50 such questions are asked each class period and all students participate at least once during a given week. The presentation will demonstrate the method and discuss the statistically significant improvement in both student learning and teacher evaluations. Attendees should leave with sufficient knowledge to attempt this new teaching style. n MA28 North Bldg 221A Maximizing Railway Network Capacity Sponsored: Railway Applications Sponsored Session Chair: Carl D. Van Dyke, TransNetOpt, West Windsor, NJ, 08550, United States 1 - Rethinking Train Scheduling to Improve Network Capacity Management Carl D. Van Dyke, TransNetOpt, 6 Snowbird Court, West Windsor, NJ, 08550, United States

129